[PDF](+66👁️) Télécharger NASA LCLUC Program: An Integrated Forest Monitoring System for Central Africa pdf

NASA LCLUC Program
An Integrated Forest Monitoring System for Central Africa



Thursday, May 20, 2004 niaporte







Final Report April 2003-2004



Principal Investigator

• Eh-. Nadine Laporte

The Woods Hole Research Center
e-mail: nlaporte@whrc.org

Main Collaborators (full list of collaborators provided at the www.whrc.org/africa)

• Dr. Jacqueline Le Moigne

NASA-GSFC, Applied Information Sciences Branch, Greenbelt, MD 20771, USA
e-mail: lemoigne@backserv.gsfc.nasa.gov

• Dr. Paul Elkan

Wildlife Conservation Society, Kabo, Rep.of Congo
e-mail: pelkan@wcs.org

• Olivier Desmet & Dominique Paget

Forest Management Unit-CIB, Pokola, Rep.of Congo
e-mail: olivier.desmet@cibpokola.com

• Dr. Andrew Pumptre

Wildlife Conservation Society, Albertine Program, Uganda
e-mail: aplumptre@wcs.org

• Patrice Gouala

Ministry of Forest, Brazzaville, Rep. of Congo,
e-mail: pgouala@yahoo.ft"

• Dr. Miro Honzack
Conservation International
e-mail: nihonzak@yahoo.com

• Fiona Maisels

Wildlife Conservation Society, WCS Gabon
e-mail: fmaisels@wcs.org



Thursday, May 20, 2004 niaport



NASA LCLUC Program
An Integrated Forest Monitoring System for Central Africa



TABLE OF CONTENTS

ABSTRACT 3

MAPPING AND MONITORING CENTRAL AFRICAN RAINFOREST 3

GOALS / SCIENCE IMPLICATION 3

APPROACH / METHODS 4

ACHIEVEMENTS /RESULTS - 4

A- DEVELOPPING AN OPERATIONAL LAND USE CHANGE MONITORING SYSTEM 4

A-1. Mapping logging history using Landsat (1970 - 2002) 5

A-2. Assessing rates of deforestation in logging towns 6

A-3. Mapping logging impact 6

B- Land COVER MAPPING FOR FOREST CONSERVATION 7

B-]. Mapping Habitat in Northern Congo 7

B-2. Mapping forest habitat loss in the AJberline region 8

C- Data FUSION & RADAR BIOMASS ESTIMATES 9

C-1- Assessing biomass using radar imagery (JERS-1 100m mosaics) 9

C-2. Advance in Data integration/ fusion -Fusion ofSAR and optical data for vegetation mapping 10

D-GOFC IN CENTRAL Africa- The OSFAC NETWORK 10

Dl -Outreach/ capacity building 10

D- 2. Image Data Acquisition and distribution in Central Africa 11

CONCLUSIONS & ISSUES 11

ACKNOWLEDGMENTS 12

PEER REVIEWED PAPERS AND BOOKS 12

REPORT AND NGN PEER REVIEWED PAPERS & CONFERENCE POSTERS 13



Thursday, May 20, 2004 niaporte



NASA LCLUC Program
An Integrated Forest Monitoring System for Central Africa

Abstract

Central Africa has the second largest unfragemented block of tropical rain forest in the
world; it is also one of the largest carbon and biodiversity reservoirs. With nearly one-third of
the forest currently allocated for logging, the region is poised to undergo extensive land-use
change. Through the mapping of the forests, our Integrated Forest Monitoring System for
Central Africa (INFORMS) project aims to monitor habitat alteration, support biodiversity
conservation, and promote better land-use planning and forest management. Designed as an
interdisciplinary project, its goal is to integrate data acquired from satellites with field
observations from forest inventories, wildlife surveys, and socio-economic studies to map and
monitor forest resources. This project also emphasizes on collaboration and coordination with
international, regional, national, and local partners— including non-profit, govenmiental, and
commercial sectors. This project has been focused on developing remote sensing products for the
needs of forest conservation and management, insuring that research findings are incorporated in
forest management plans at the national level.

The societal impact of INFORMS can be also appreciated through the development of a
regional remote sensing network in central Africa. With a regional office in Kinshasa,
(www.OSFAC.org) , the contribution to the developemnt of forest management plans for 1. 5
million hectares of forests in northem Republic of Congo (www.tt-timber.com) , and the
monitoring of park encroachments in the Albertine region (Uganda and DRC)
(wvvw.albertinerift.org) .

Key words:

Research Fields: Habitat Conversion, Deforestation, Vegetation Mapping
Geographic Area/Biome: Central AfHca, Tropical Forest
Remote Sensing: Radar, MODIS, IKONOS, LANDSAT, digital video surveys
Methods/scales: Local to Regional Scale, Data Fusion, Change detection

Mapping and Monitoring Central African Rainforest

Remote sensing research and applications (75%), Social science (25%)
Carbon (25%), GOFC (50%), Biodiversity Monitoring (25%)

Central African forests are one of the largest carbon reservoirs on Earth but relatively little is
known about the impact of agriculture and logging on carbon stocks in this region. Changes in
forest biomass under different land-use scenarios have been addressed in other tropical forest
regions (e.g., the Amazon Basin), but the modes of forest harvest and use are very different in
Afhca. Levels of forest fragmentation and the intensity of forest biomass removal is still largely
imknown or poorly docimiented. As part of INFORMS we developed a strategy to address these
needs, and made progress on implementing them, as surcimarized below.



Goals / Science implication

• Better characterisation of tropical forest land surfaces and processes.

• Develop multi-scale multi-sensor data integration methods and appropriate validation tools.

• Develop o|>erational monitoring systems to support environmental policies



Thursday, May 20, 2004 niapor



NASA LCLUC Program
An Integrated Forest Monitoring System for Central Africa

• Integration of central African research scentists in regional science activities (eg GOFC-
OSFAC).

Approach / Methods

Develop new forest monitoring approaches under the framework of Global Observations of
Forest Cover (GOFC), including:

• Development of methodologies to assess and map central Africa forests and rates of
deforestation using multi-sensor, multi-scale satellite observations, providing improved
vegetation maps for applications at the local and regional scale.

• Development of forest monitoring techniques integrating new remote sensing information,
biodiversity and forestry information in collaboration with international organizations and
local stakeholders to facilitate forest conservation.

In this brief final report we will focus on our achievements. More details on approaches and
methods can be found at www.whrc.org/afHca .

Achievements / Results

Characterization and mapping of land cover/land use in Central African rainforest is complex.
This complexity is exacerbated by (1) the diversity of human land uses and (2) the lack of full
and continuous cloud-free coverage by any single remote sensing instrument. In order to provide
improved vegetation maps of Central Africa and to develop forest monitoring techniques for
applications at the local and regional scales, and develop remote sensing capacity we have
focused on:

A. Developing operational land-use monitoring systems

B. Developing methods for land-cover mapping for forest conservation

C. Assessing the use of Radar imagery for biomass mapping and data fusion

D. Developing a remote sensing network as part of the GOFC initiative



A. Developing an operational land-use change monitoring system

Selective logging is the most prevalent land use in the Congo Basin (30% of the forest)
and more timber companies are expected to begin operations in the coming year (see figure 1).
Although historically logging intensity is relatively low in most of Central Afiica compared to
Southeast Asia and the Brazilian Amazon, harvesting volumes are steadily increasing. Even
long established timber companies are exploring secondary species as parts of the forest are
exhausted of primary timber species. Over-harvesting of timber is rarely monitored or controlled,
and long-term forest ecosystem, community, and population responses to this high-grading
process are largely understudied. Furthermore, logging roads create access to commercial
hunters and poachers into previously remote areas of the forest, and migration into logging towns
causes increasing rates of deforestation.

The NASA-INFORMS project produced crucial information for the forest management
of northern Republic of Congo. These results include 1) monitoring the increase of logging roads
in northern Congo ; ) assessment of deforestation rates around major population centers (logging
towns), and 3) the development of a forest logging intensity index



Thursday, May 20, 2004 niaporte



NASA LCLUC Program
An Integrated Forest Monitoring System for Central Africa




Latest timber concession limits ■■ Protected areas | ILandsat historical logging monitoring (400x 400 km]



Figure 1. Distribution of timber concessions and protected areas in Central Africa.

A-1. Mapping logging history using Landsat (1970 - 2002)

In the northern Republic of Congo alone, the total length of logging roads established in
the last 30 years was estimated to be more than 6,000 km, or two times the total length of
primary roads in the entire country (Figure 2). This equates to an eleven-fold increase in the last
25 years. Similar rates of logging expansion are anticipated in D.R.C. when the political
situation is stabilized.



1970s
S9€km






mid-1980s
1760 ton



^ 2000-2001
"^ 5520 km






i^^ 2002 3>^

\ 6846 km A ^



r_




T* /■




~ 150 km per year > 650 km per year

Figure 2. Historical logging monitoring in Northern Congo: mapping roads from Landsat
images of 1970s-2000s.



Thursday, May 20, 2004 niapor



NASA LCLUC Program
An Integrated Forest Monitoring System for Central Africa

References for results:

Laporte, N.T., and T.S. Lin. 2003. Monitoring logging in the tropical forest of Republic of Congo with Landsat

imagery. International Geoscience and Remote Sensing Symposium (IGARSS), July 2003 Toulouse, Ref #

8.1812- INT-A17. http://www.igarss03.com/
Laporte, N., T.S. Lin, and D. Devers. 2003. Impacts of Large-Scale Selective Logging on Ecosystem Services in the

Northern Republic of Congo, AGU Chapman Conference Ecosystem Interactions with Land Use Change 14-18,

June 2003 Santa Fe, New Mexico.

A-2. Assessing rates of deforestation in logging towns



Northern Republic of Congo is an important area to develop an understanding of the
impact of logging on central African forests. Logging is the dominant land use for most of this
otherwise largely undisturbed natural region (and National Park). Current rates of deforestation
associated with logging in the region are poorly documented. Our mapping of land-use change
associated with logging at Pokola shows that logging towns have a higher rate of deforestation
compared to non-logging towns (Ouesso). In logging towns new fields are cleared in forests,
whereas in non-logging towns clearing takes place primarily in fallow areas which are easier to
clear and more accessible. Urbanization is also progessing rapidly in logging towns. Figure 3
illustrates deforestation around the logging town of Pokola in Northern Congo.



Pokola 1990



Pokola 2001



Change Detection
1990-2001




Land Cover Map Legend

JBI Bare soil

^i Agriculture/ Degraded

HM Forest

■■ Water



Land Cover Change Legend
^B No Change
^H Deforestation
BM Urbanization
■■ Water



-andsat-derived maps Windows size: 7 kmx JOkm

Figure 3. Mapping land cover change in logging town (Pokola).



References for results:

Laporte, N. T, T. S. Lin, J. LeMoigne, D. Devers, and M. Honzak. Toward an Integrated Forest Monitoring System
for Central Africa. In: Land Change Science: Observation, Monitoring, and Understanding Trajectories of
Change on the Earth Surface, Ed. G. Gutman. NASA-LCLUC Program, in press.

A-3. Mapping logging impact



Millions of hectares of forested land in the region are under concession (i.e., allocated for
logging) - Figure 1 . We have developed a simple "logging index" derived from Landsat satellite
imagery. It allows estimating the number of trees harvested by forest unit (Figure 4). The
development of such indicators is important because logging intensity is likely to intensify in the
region (as more species are harvested per unit area). Furthermore, the more trees that are
removed, the larger the impact on carbon sequestration and fluxes.



Thursday, May 20, 2004 niaporte



NASA LCLUC Program
An Integrated Forest Monitoring System for Central Africa



R^ = 0.5036




20 30 40 50 60

Number of trees harvested per 50-ha parcel)



Figure 4: Developing a logging forest index for central Africa

References for results:

Laf)orte, N. T, T. S. Lin, J. LeMoigne, D. Devers, and M. Honzak. Toward an Integrated Forest Monitoring System
for Central Africa. In; Land Change Science: Observation, Monitoring, and Understanding Trajectories of
Change on the Earth Surface, Ed. G. Gutman. NASA-LCLUC Program, in press.

Lap>orte N., 2003, Utilisation de la teledetection pour la gestion des ecosysteme forestier du nord Congo- Cas des
UFA Kabo, Pokola et Loundoungou, Rapport Technique au Wildlife Conservation Society, New York, USA.



B. Land cover mapping for forest conservation



Biodiversity conservation depends strongly on the management of both protected areas and
their buffer zones. Unfortunately, in most of Central Africa, very little is known about the
distribution of vegetation types and associated threats, such as deforestation, forest degradation,
and forest fragmentation. While forest covers roughly 45% of Central Africa, only 10%, or
1 80,000 km^, is currently protected.

The need for forest monitoring in Central Africa is urgent. National institutions in the region
lack the most basic information to make land-use decisions and policy. Recent vegetation maps
produced for the region were generated from low-resolution satellite imagery including AVHRR,
SPOT vegetation and MODIS. These coarse-scale maps are useful for monitoring land cover at
the regional level, but they are poorly adapted to forest conservation needs at the landscape level.

B-1. Mapping Habitat in Northern Congo

Wildlife management requires knowledge of habitat distribution and potential threats.
Landsat TM imagery allows us to monitor forest cover around parks. A vegetation map of the
Nouabale Ndoki reserve( Northern Congo) was produced in collaboration with WCS researchers.



Thursday, May 20, 2004 niaport



NASA LCLUC Program
An Integrated Forest Monitoring System for Central Africa




Monodominant evergreen Gilbertiodendron dewevrei forest
Mixed species semi-evergreen forest

closed canopy

semi-open canopy, type I

semi-open canopy, type 2

open canopy

degraded forest / fallow

very open canopy with Marantaceae dominated understory
Mixed species wetland

Closed canopy swamp

Riparian forest & Raphia swamp

Open canopy swamp & Cyperaceae marshopen canopy



! Savana



Agriculture



I Bare soil



I Water



Figure 5. Detailed vegetation mapping in the northern Congo using Landsat (14 of the 18 classes are shown here)



References for results:

Laporte N., 2003. Utilisation de la teledetection pour la gestion des ecosystemes forestiers du nord Congo- Cos des
UFA Kabo, Pokola et Loundoungou, Rapport Technique au Wildlife Conservation Society, New York, USA.

Laporte, N. T, T. S. Lin, J. LeMoigne, D. Devers, and M. Honzak. 2004Toward an Integrated Forest Monitoring
System for Central Africa. In: Land Change Science: Observation, Monitoring, and Understanding
Trajectories of Change on the Earth Surface, Ed. G. Gutman. NASA-LCLUC Program, in press.

Lin, T.S., and N.T. Laporte. 2003. Using Aerial Videography to Validate Land Cover Map in the Tropical Rain
Forest of Central Africa. Poster for American Geographic Union (AGU) 2003 Fall Meeting, 8-12 Dec, San
Francisco.

B-2. Mapping forest habitat loss in the Albertine region

Because of the fragmented park system in south West Uganda and their lack of
connecting corridors, the sustainabiHty of viable wildhfe populations is under severe threat. We
found the Albertine forest in general to be fragmented, even in the Democratic Republic of
Congo. In SW Uganda, dense forest can be found only in protected areas. To date rates of
deforestation in this region have been poorly documented, and most deforestation occurs within a
short distance from protected areas. The largest deforestated areas were located in the northern
part of the protected area network, around Bugoma, Budonga, and Kagombe forests.





Figure 6a, b, c: Budongo Landsat TM deforestation map a) , MODIS tree cover b), and population density map c)



Thursday, May 20, 2004 niaporte



NASA LCLUC Program
An Integrated Forest Monitoring System for Central Africa



References for results:

Plumptre, A.J., N. Laporte, and D. Devers. 2003. The Biodiversity of the Albertine Rift - Threats to sites, in The

Biodiversity of the Albertine Rift - Albertine Rift Technical Reports III - Edited by A.J. Plumptre, et ai..

Chapter 9, pp 77-82.
Laporte N., and T. Lin. 2003. Land Use Land Cover Change in the Albertine Region of Uganda. Paper Number:

B42A-0943- American Geophysical Union-2003 Fall Meeting, 8-12 December 2003, San Francisco, California,

USA.



C. Data fusion & radar biomass estimates

C-1- Assessing biomass using radar imagery (JERS-1 100m mosaics)

We explored the utility of 100-m JERS-1 radar imagery for above-ground biomass
estimation across a range of 1-ha plots in 61 study sites throughout southern Cameroon. Many
studies have shown a positive correlation of radar backscatter to total above-ground biomass of
different forests in the Northem Hemisphere. But tropical forests have a more complex structure
and higher biomass than the northem forests, and it is expected that the assessment of south
Cameroon forest biomass, using JERS-1 100 m mosaics provided by JPL, will be more
challenging.

The field biomass measurements, collected in 1995 by Honzack (Co-I), were compared
with the normalized backscatter of the 1996 JERS-1 mosaic produced by the NASA Jet
Propulsion Laboratory (http://trfic.ipl.nasa.gov) . We found poor relationships between the
above-ground biomass measurements and backscatter for both low and high water mosaics.
These findings suggest limited utility of JERS-1 radar imagery for biomass estimation across
tropical Africa. Similar limitations have been noted in the Amazonian forests (Salas et al.,
2002).

Figure 7a,b Mean above ground biomass v. JERS-1 Normalized backscatter (Sigma 0) in Cameroon



-6.5



O -7

2,

o

fS

E
;^-7.5



November 1996 (wet season)



^ 5&



10O - ia> ■ ■' 200

y = 0.0074Ui{x)- 7.^73
- R^^O.OOOt




February 1996 (dry season)



-6.5



Mean above ground biomass (Mg / ha)




Mean above ground biomass (Mg / ha)



Thursday, May 20, 2004 niapor



NASA LCLUC Program
An Integrated Forest Monitoring System for Central Africa

References for results:

Laporte, N. T, T. S. Lin, J. LeMoigne, D. Devers, and M. Honzak. 2004 Toward an Integrated Forest Monitoring
System for Central Africa. In: Land Change Science: Observation, Monitoring, and Understanding
Trajectories of Change on the Earth Surface, Ed. G. Gutman. NASA-LCLUC Program, in press.



C-2. Advance in Data integration/fusion -Fusion of SAR and optical data for vegetation
mapping

A wavelet-based fusion method was developed for integrating high-frequency components of
higher spatial resolution data (SAR data at 6m resolution) and low-frequency components of
lower spatial resolution data (Landsat-TM at 30m resolution). The fusion provides a new image
data set at 6m spatial resolution, which contains more detailed texture features used to improve
land cover classification. At the same time, the fusion preserves the large homogeneous regions
that are observed by the Thematic Mapper sensor.

References for results:

Le Moigne, J., N. Laporte, and N.S. Netanyahu. 2001 . Enhancement of tropical land cover mapping with wavelet-
based fusion and unsupervised clustering of SAR and Landsat Image Data, International Society for Optical
Engineering (SPIE), Toulouse, 17-21 September 2001, Toulouse, France, 6p.



D- GOFC in central Africa - The OSFAC network

International efforts are underway to improve operational forest monitoring. At the first
Central Africa "Global Observation of Forest Cover" workshop held in February 2000 in
Libreville, Gabon, several issues limiting the development of operational forest monitoring
systems were identified by national forest services and their international partners. These
limitations included lack of technical and financial resources, poor access to data and
information (including the Intemet), and lack of training facilities and opportunities. It was also
unanimously recognized that remote sensing is a key component to any forest monitoring system
and that, at each stage of scientific understanding, there must be transfer of relevant information
to policy and decision makers.

D-1. Outreach/ capacity building

In collaboration with the TREES project, we organized GOFC/CARPE regional forest
monitoring workshops with international NGOs and national forest services. The first workshop
took place in Libeville,Gabon, in the spring of 2000 and was followed by a more technical
workshop in Lope, Gabon in July 2000. The objective was to build a national capacity for
operational forest monitoring. More than 30 in-country researchers participated in the workshop,
which focused on practical applications of satellite data to forest monitoring and conservation.
For 2003-2004, CARPE/USAID funds have been secured for the Central Africa GOFC Regional
Office in Kinshasa (OSFAC).

References for results:

Laporte N. 2000. Les Donnees basse resolution pour le suivi de la dynamique de la vegetation tropicale. Atelier
GOFC (Global Observation of Forest Cover) Afrique Centrale, Libreville 22-24 Fevrier2000, lip.



Thursday, May 20, 2004 niaporte



NASA LCLUC Program
An Integrated Forest Monitoring System for Central Africa

Laporte N. 2001. Tropical deforestation in Ceittral Africa. Brie fing sheets #6, Bicxiiversity Support Program, March

2001 ,Washington E>C, 4p.
Laporte N. 2001. Forest monitoring in Central Africa. Briefing sheets #13, Biodiversity Support Program, March
2001, Washington DC, 6p.



D-2. Image Data Acquisition and distribution in Central Africa

Following the recommendation of the 200 GFC meeting, a series of Landsat TM, ETM+
and IKONOS images have been acquired through the NASA data purchase program and
distributed to in-country CARPE and GOFC collaborators in order to implement local-scale land
cover mapping activities. The remote sensing data acquired as part of this project and the NASA
data buy have been distributed to our collaborators in the region. They are also freely available
from GLCF and the OSFAC website (www.osfac.org ). This regional network allows remote
sensing information of this region to be shared. However, Internet access is still a limiting factor
in central Africa in terms of speed and cost, and most of the data sets are still distributed via CD.

Conclusions & Issues

This NASA fimded project has advanced knowledge of land-cover and land-use change
in Central Africa and has promoted forest management and monitoring techniques using remote
sensing. Capacity building in the use of remote sensing was built through organized workshops,
scientific collaborations and data sharing with national forest services and conservation
organizations, and through integration with the GOFC-OSFAC network. Nonetheless, Central
Africa remains greatly imderstudied, partly due to diverse cultural and language differences and
poorly developed infrastructure, but primarily due to the difficulty of working in such politically
unstable countries. Since the mid-1990s, this "hot spot" of biological diversity and enormous
carbon storage has undergone a series of devastating ethnic conflicts, which are to blame for the
death of more than 3 inillion and the displacement of 2.7 million people. Partially because of
this displacement, localized increases in forest loss are evident throughout the region, caused by
agricultural expansion, charcoal production, and livestock grazing. Funding by NASA for this
project has contributed significantly in improving forest management through the provision of
low cost imagery, GIS data sets, and scientific results, but perhaps most importantly through
capacity building and the promotion of international research collaborations. One notable social
and cultural success of the study has been the creation of remote sensing/GIS units in each of our
primary study sites, including at a large private logging company (CIB) interested in developing
reduced impact logging techniques and achieving its timber certification.

Building upon these advances and coupling remote sensing science with specific
information on ecosystems structure and function are the only ways to continue to track the rapid
changes occurring in this vast forested region (> 1.8 million sq km), and to predict the impact of
these changes on human populations, biodiversity, and carbon pools and fluxes. Little
LANDSAT imagery after 2003 is available for the region, and MODIS does not provide the fine
scale information needed for forest management by the logging companies, the national forest
services, or the implementation of the forest monitoring system. Consequently, we recommend
continued NASA support for the acquisition and processing of Landsat imagery of the regionin
order to attainclear images for this extremely cloudy region.



Thursday, May 20, 2004 niaport



NASA LCLUC Program
An Integrated Forest Monitoring System for Central Africa

Results from our research have been focused on the development of an operational forest
monitoring system in the region, and in this sense the project has been extremely successful.
Results have been presented at numerous national and international conferences, distributed to
in-country Forest Services, NGOs, and the collaborating timber products company. Some of the
results have been published in scientific journals, and there will be further publications as we
document and extend our research and applications in the region. We hope to continue to
contribute to the NASA programs as part of the Land Cover/Land Use Change Science Team,
and by advancing our research activities in the region through the NASA Carbon Cycle Science
and related programs. We will also continue to pursue our remote sensing capacity building and
forest monitoring activities in the region, where possible, under the GOFC-OSFAC and CARPE-
USAID umbrella.



Acknowledgments

This research was supported by the NASA Land Cover Land Use Change Program and the
Central AfHca Regional Program for the Enviroimient (CARPE). It would not have been
possible without the logistic support of the Wildlife Conservation Society (Paul Elkan and Fiona
Maisels) and the Congolaise Industrielle des Bois (Olivier Desmet, Dominique Paget). We want
to thank the following individuals for generous donation of their time and their work dedicated to
preserving these unique ecosystems: Steve Blake, Brian Curran, Bourges Djoni-Djimbi, Sarah
Elkan, Mike Fay, Steve Gulick, Richard Malonga, Antoine Moukassa, and Chris Wilks. We also
thank Yves Dubois, Leon Embon, Patrick Geffroy, Frederic Glaimaz, Jackie Glaruiaz, Patrice
Gouala, Christian Guyonvaro, Gregoire Kossa, Tom van Loon, Alfred Tira, and Luca van Der
Walt for field assistance and working toward the sustainable use of the region's forest resources.
I want also to thank my former research assitant at the University of Maryland, Didier Devers,
for helping with the development of the OSFAC network and his technical assistance, and Debra
Fischman and Jeremy Goetz for assisting in the mapping of logging roads.



Peer reviewed papers and Books

•Chan J. C.W., Laporte N., Defnes R., (2003), Texture classification of logging in tropical Africa
using machine learning algorithms. International Journal of Remote Sensing, 24(6): 1401-1407.

•Laporte, N. T, T. S. Lin, J. LeMoigne, D. Devers, and M. Honzak, Toward an Integrated Forest
Monitoring System for Central Africa. In: Land Change Science: Observation, Monitoring, and
Understanding Trajectories of Change on the Earth Surface, Ed. G. Gutman. NASA-LCLUC
Program, in press.

•Plumptre, A.J., Laporte, N. and Devers, D. (2003) The Biodiversity of the Albertine Rift -
Threats to sites, in The Biodiversity of the Albertine Rift - Albertine Rift Technical Reports III -
Edited by Plumptre, A.J., Behangana, M., Davenport, T.R.B., Kahindo, C, Kityo, R., Ndomba,
E., Nkuutu, D., Owiunji, I., Ssegawa, P. and Eilu, G., Chapter 9, pp 77-82,



Thursday, May 20, 2004 nlaporte



NASA LCLUC Program
An Integrated Forest Monitoring System for Central Africa



•Wilkie D., Laporte N. (2001), Forest area and deforestation in Central Africa: Current
knowledge and future directions, in W. Weber, A. Vedder, S. Morland, L. White (Eds), African
Rainforest Ecology and Conservation, Yale University Press, 1 19-1 3 8p.




Report and non peer reviewed papers & conference posters

are posted on our INFORMS - NASA/LCLUC website at:



http;//www. whrc.org/africa
(Starting June 2004)



http://luci.umd.edu/tcluc/
(Before June 2004)






Lire la suite

• Taille
  15.71 MB
• Nombre de pages
  15